Spyder 4G64 Spyder GS Fuel Pump Relay Module Bypass

Disclaimer: This documents mods made by me. I make no claims about any mods you make on your car that follows my process.

Hours were spent trying to trace the cause of our performance problems on our 1999 4G64 Spyder GS. The car would run great, switching from open loop to closed loop, until it warmed up. Then it would stumble and die. Since the problem was happening starting at different temperatures, it was not easy to pinpoint. Once, when I went to unplug the OBDII logging cable, I about burned my hand on the Fuel Pump Relay Module, which has a metal case. I knew something was wrong.

In 1999 (and 1993-97 for Galant), Mitsu added a Module to control the Fuel Pump Relay. We have been unable to determine the need for it, but some quick research indicates it was an attempt to vary the voltage to the Fuel Pump Relay in order to throttle the speed of the Fuel Pump.

Here is what the Fuel Pump Relay Module looks like, with the cover removed

You must be logged in to view this image or video.

Here is how the fuel circuit is wired in the 1999 4G64

You must be logged in to view this image or video.

Complicated. By unplugging the Fuel Pump Relay Module, and making three connections, you can eliminate the need for the Module, and just use the Fuel Pump Relay, like all other DSMs.

You must be logged in to view this image or video.

Your Fuel Pump Check Connector will still work. Keep in mind that the Fuel Pump will now run with the key in the "ON" position.

If you are going to upgrade to a GST or GSX ECU, keep in mind that Pin #8 on those ECU's is a switched ground, so you could make your new grounding connection (#4) on Pin #8, and the Fuel Pump would only run when the engine is cranking.

Hopefully this helps save hours of time for someone who is having mystery rough-running, or no-start issues on their 4G64 GS.

 

[Hristo]

The answer hasn’t changed: You have to use a volt meter to check to make sure your Pin #8 on your ECU goes to ground when the engine is cranking and running.

No electric pump will run without a good ground connection.

I will check that, but what if I don't have ground on pin 8?

 

[DogWhistle]

I will check that, but what if I don't have ground on pin 8?

Then you either are not running a ‘95-‘96 ECU, or your ECU has a burnt ground circuit.

Your only choice at that point would be to either wire in a switch for grounding — if you don’t want the pump to start with the key in the “Run” position — that you can flip on just prior to starting the engine.

 

[Potmus63t]

Nothing about eliminating the fuel module involves changing any wiring at the pump. You have your pump rewired using a relay for the battery power switched by the 12v fuel pump feed. When we did our fuel pump rewire, we left the fuel pump ground traveling in the harness up to the front fuel module.

Do you mind explaining this a little more? I am in the process of doing the fuel pump rewire mod, and then I was going to go after the relay bypass as mentioned in your original post (from over a decade ago!).

All the rewire write ups I’ve come across show the same diagram as pictured below. It sounds like you’re saying you didn’t touch the ground wires at the pump, and I’m just curious of how you wired it compared to the diagram shown. I may just be confusing myself because I’m not at my car right now looking at it, but I’m trying to figure it out so I can do it right, and do it once.

You must be logged in to view this image or video.

 

[Potmus63t]

It appears that the OP has his fuel pump wired as STM suggests. It’s slightly different from how many of the threads on the forums suggest with the diagram that’s been floating around. Here, I crudely drew out the differences. It appears the way to go is the ‘STM’ method if you’re using a turbo ECU which would allow the grounding needed at the front while cranking, by connecting pins 4&2 (B-W & W-R).

You must be logged in to view this image or video.

If this plug was rotated a bit, it would better illustrate what is going on in regards to the jumpers installed onto the plug and how it can be replicated.

You must be logged in to view this image or video.

Pins 1&3 are jumpered for power (B/R & B/L).

-THEN-

If using 4g64 ECU, pins 4&5 are jumpered (B/W & B).
If using 4g63 ECU, pins 4&2 are jumpered (B/W & W/R).

 

[steve]

It helps to have a diagram of the common Bosch relay when looking at suggested wiring schematics.

You must be logged in to view this image or video.

The only difference between the STM and the other is how they draw the ground connections. Note that in both cases the 86 terminal of the relays coil is connected to the + supply (usually from the original Fuel Pump Relay output) and the 85 terminal to the negative or ground return which are the same. And the switched N.O. (Normally Open) terminal 87 runs to the positive side of the fuel pump. The common terminal 30 of the relay is used as the input and connected to the battery positive (hopefully fused back close to the battery).

The last detail is the fuel pump ground. Now the whole point of the Fuel Pump Relay mod is to eliminate the normal voltage drop that happens when you try and draw lots of current through long little wires so I submit that after you run nice "fat" wire to the relay you want an equally "fat" ground and rather than run one back to the battery or use the existing wire only to connect it to a smaller one at the connector (pins 4 and 5 here on a Spyder GS), you use the car body close to the pump.

Since this is in a thread about bypassing the Spyder GS Fuel Pump Relay Module and not a thread on generic fuel pump relay mods, I'll also point out that on the GS you really should bypass this module before you ever try and do the fuel pump relay mod. You want to make sure that the trigger (the wire to the 86 terminal) isn't being controlled by whatever this module is putting out to the pump. It's likely to be a varying voltage to adjust the pump speed and not what a relay wants to see.

Lastly, assuming your relay doesn't have a diode in parallel with the coil the polarity of how you connect to terminals 85 and 86 doesn't matter. (ie: which one is connected to ground.) If there is a flyback diode you need to do it correctly and I'll leave that as an exercise for the students.

 

[Potmus63t]

1998 Eclipse Spyder GS
95 Turbo ECU w/ ECMLink

I did a fuel pump swap before I started the engine build (63 head/turbo) to verify the new walbro 190 was working correctly. It was. Now I went ahead and installed the STM fuel pump rewire kit (relay/wiring) and wired it up how STM suggested.

In looking into how to bypass the additional spyder 'modulating' relay, I found 2 versions. One is the version in the ‘how to add a turbo ECU to your spyder gs’. This involved jumpering pins 1&3 (B/R & B/L) for power, and then for pin 2 (R/W) the tutorial said to run it to the fuel pump relay(?) on the passenger side of the car and to cut the solid black wire (pin 2) going into the fuel pump relay, and attach the connector side of this wire to the R&W wire. I initially tried this, and the fuel pump wouldn’t run while cranking, nor when activating it through ecmlink.

So I put the wire back together on the passenger side, and wired pin 4&2 ( B/W & W/R) together that went to the relay i am eliminating. This allowed the fuel pump to start via ecmlink when checked ‘always on’. I was able to start the car and it ran for about 5 seconds at high idle (2000rpm) on a cold start (45* F ambient temp), which I summed up as normal, engine sounded good, THEN it died.

Now I’m able to crank the engine, but the relays (dash, and new relay at the fuel pump) no longer click. When trying to run fuel pump from ECMLink, the relays don't click. I verified the fuse from the battery is good, but it seems I’m having a power issue at the relay(s). All I find on the forums is possibly the MPI relay isn’t working, and is somewhat commonly a source of problems.

Edit-

So I went over all the wiring I did, and even respliced connections to be sure. Connections are all good. Was able to hear the pump prime when hooking power to the ‘pump check’ plug along the firewall. So I read it’s either mpi/fuel relays, or possibly ignition switch not sending power to the mpi relay. I did pull the mpi relay and checked resistance across the coils and got a good reading. When putting power to it, it does click, but I can’t hold power to it and test at the same time…just a pain in the ass.

Tomorrow I’ll check the fuel relay, and see what that does. I’ll inspect under dash wiring and make sure the harness is solid in the ECU, but just thought I’d share an update.

 

[Potmus63t]

1998 Eclipse Spyder GS
95 Turbo ECU w/ ECMLink

I did a fuel pump swap before I started the engine build (63 head/turbo) to verify the new walbro 190 was working correctly. It was. Now I went ahead and installed the STM fuel pump rewire kit (relay/wiring) and wired it up how STM suggested.

In looking into how to bypass the additional spyder 'modulating' relay, I found 2 versions. One is the version in the ‘how to add a turbo ECU to your spyder gs’. This involved jumpering pins 1&3 (B/R & B/L) for power, and then for pin 2 (R/W) the tutorial said to run it to the fuel pump relay(?) on the passenger side of the car and to cut the solid black wire (pin 2) going into the fuel pump relay, and attach the connector side of this wire to the R&W wire. I initially tried this, and the fuel pump wouldn’t run while cranking, nor when activating it through ecmlink.

So I put the wire back together on the passenger side, and wired pin 4&2 ( B/W & W/R) together that went to the relay i am eliminating. This allowed the fuel pump to start via ecmlink when checked ‘always on’. I was able to start the car and it ran for about 5 seconds at high idle (2000rpm) on a cold start (45* F ambient temp), which I summed up as normal, engine sounded good, THEN it died.

Now I’m able to crank the engine, but the relays (dash, and new relay at the fuel pump) no longer click. When trying to run fuel pump from ECMLink, the relays don't click. I verified the fuse from the battery is good, but it seems I’m having a power issue at the relay(s). All I find on the forums is possibly the MPI relay isn’t working, and is somewhat commonly a source of problems.

Edit-

So I went over all the wiring I did, and even respliced connections to be sure. Connections are all good. Was able to hear the pump prime when hooking power to the ‘pump check’ plug along the firewall. So I read it’s either mpi/fuel relays, or possibly ignition switch not sending power to the mpi relay. I did pull the mpi relay and checked resistance across the coils and got a good reading. When putting power to it, it does click, but I can’t hold power to it and test at the same time…just a pain in the ass.

Tomorrow I’ll check the fuel relay, and see what that does. I’ll inspect under dash wiring and make sure the harness is solid in the ECU, but just thought I’d share an update.

So I’ve tested both MPI and fuel relays, and also the new relay for the pump rewire. Everything checked out good.

Im going to hook everything back up, and try as Steve recommended, and run my ground directly to chassis at the fuel pump instead of relying on the ground that runs back up front. This SHOULD provide a better ground for the pump. After this, if I still have an issue, I’ll verify whether or not my new pump is still working or if it somehow prematurely failed by powering it directly.

 

[Potmus63t]

Final update.

Ran through this thread again…what I found is my ECU wasn’t providing ground at pin 8 like it should. Should have checked that first with how many times it was mentioned in here. Wound up running the B/W wire at the relay module to ground and sure enough, pump started up with key on, and car started and stayed running. I just wanted to provide an update here for any future lookers.

Side note, gotta adjust my exhaust as it vibrating on chassis in a spot, but after almost a year of swapping my spyder gs into a turbo car with 63 dohc swap, it’s alive and ready to take the next steps.